Hacker News new | past | comments | ask | show | jobs | submit login
AmazonBasics Electronics Fire Safety Investigation (cnn.com)
98 points by SethMurphy 39 days ago | hide | past | favorite | 20 comments



Shopping with Amazon has become increasingly difficult. I can no longer rely on the reviews, and the volume of results that come up for even a simple search make for a lot of searching. Amazon really, really need to stop being the dumpster for cheap and nasty China-manufactured goods. And yes, of course I'm welcome to shop elsewhere if I don't like what's on offer, and that's what I'm increasingly doing. Bezos' loss, but he'll never notice.


I miss the Amazon from yesteryear when Prime was first launched. You could confidently read and trust the reviews, 1-click buy the item when satisfied, and have it show up 1-2 days later. Today that magic is lost; it’s mostly a huge junk store with untrustworthy reviews.

Something I can’t really get my head around is that it seems like such an obvious problem to identify within Amazon’s ranks, and correcting it feels like something within their means, but they just ... don’t.

Are there any e-commerce sites that exist today with the same magic Amazon use to have?


It appears to be a similar problem akin to facebook, twitter etc. which promote "every" content no matter how bad (e.g. trolling etc.) it is for as long as the underlying metrics (e.g. click rates, revenue) appear to go up. However, they don't see the second order effect, namely that customers lose interest in engaging with the sites over the long run and start looking for alternatives.


Articles like this aren't helpful because they pick and exhibit a few case studies instead of attempting to estimate base rates and risk proportions - of course some Amazon products have caught on fire, Amazon has sold a billion of them. What needs to be provided in this type of reporting is actual estimates of how many of the products may be dangerous, which is of course never provided or even attempted as an estimate.


I disagree with the assertion that this type of article isn't helpful. I would say it's 'not scientific', but not unhelpful.

I don't think Amazon is likely to release those statistics anytime soon, but I'm glad people are talking about this as it may change perception on both sides. Either the consumer sees this article and purchases less 'Amazon Basics' branded products, or someone inside Amazon sees it and decides to do something about it.

Not optimal, but far from unhelpful.


> Articles like this aren't helpful because they pick and exhibit a few case studies instead of attempting to estimate base rates and risk proportions

So you're saying articles aren't allowed to criticize manufacturers without proprietary information that only the manufacturer itself has?

A better question might be: Why aren't Amazon publishing the number of units sold relative to number of fires? Instead of putting the journalist trying to cover this with imperfect information into the dock, let's see Amazon defend it.


>So you're saying articles aren't allowed to

Nope, I said nothing of what is allowed, and the majority of reports like this will never contain useful statistics regardless. I just said it's not helpful to consumers, because it doesn't actually help them make a better decision when you don't know if the brand is more or less safe than the alternatives.

They could easily write an article about a huge problem someone had with product X. Then consumers instead buy product Y, which could have a higher rate of problems, and the article has done harm instead of good. I can't make a claim about this article, but given they're just written for clicks, it's not like they have the consumer's interests in mind either way.


> I just said it's not helpful to consumers, because it doesn't actually help them make a better decision when you don't know if the brand is more or less safe than the alternatives.

I agree. But I am confused why this lack of statistics is the journalist's fault instead of either the manufacturer who has is sitting on the information, or lack of regulatory rigor that could force public disclosure.

You haven't proposed how the journalist gets this proprietary information? Only that they should sit down and be quiet until they have it somehow.

If your point is that all manufacturers should disclose, so we can compare/contrast, I agree entirely. But it comes across more like you think journalists shouldn't report on problems if the people they're reporting on aren't cooperating vis-a-vis full disclosure. Which is tantamount to the end of investigative journalism.


The journalist definitely cannot access the information in the detail we'd want them to, that's true. My issue is more with their phrasing, because never would an article provide basic nuance like "This problem might not be common, and we don't know about how common it is in other products, these are just some randomly-selected cases we have found that happened in the past few years".

I think even with proper disclosures, we wouldn't necessarily be getting useful articles, but perhaps it'd at least help by enabling them to be more easily written.


They did look at a similar product from another manufacturer and compare the rate of safety concerns in reviews as a part of the article, but it was a tiny part. I would have liked to see a full table of: Amazon Product vs Multiple Similar Products though.


The articles uses Amazon's own reviews to provide statistical indications that the failure rate is higher for some Amazon Basics items:

>Since the microwave's release in the fall of 2018, its product page has been flooded with reports from consumers about problems including flames, smoke and sparks. These kinds of reviews made up roughly 5% of the AmazonBasics microwave's more than 3,000 reviews as of February [...] A microwave that has been reviewed less frequently but is the same size and wattage had only 10 reviews describing similar safety issues -- amounting to around .7% of its roughly 1,350 reviews on Amazon

and..

>Amazon said its own analysis, which added global reviews about the AmazonBasics surge protector, found 1.1% involved claims of overheating, fire and other dangers. One former AmazonBasics product manager ...[said] that a ratio of around .05% would have been seen as more acceptable when she worked there.

Not smoking guns (wires?) but certainly indications that further investigation is warranted.


I found a few parts of this article to be quite insightful, barring a statistical analysis using data that's not available:

* The use of a real lab to tear down and inspect a microwave, identifying a real design flaw leading to a safety hazard. This seems to indicate that Amazon did not send their product for independent inspection before sourcing it, or that the labs or staff they used to do so were less thorough or competent than the lab CNN found.

* Interviews with former Amazon staff indicating that standards have likely been loosened, because public available information (like reviews) which would have triggered a product's removal in the past no longer seem to do so.


The reason a news outlet can't publish actual data is that Amazon doesn't share any of it. This is the only way to force them to do so.


I was under the impression AmazonBasics was very much like Costco’s Kirkland brand, where they relabeled some other manufacturer's products. Maybe I’m wrong. Or it’s hit and miss.


This article is just a stupid hit piece.

A USB cord is not responsible for detecting short circuits. That's the job of the charger not the cord.

The best cord in the entire universe cannot prevent short circuits.


A short circuit is usually a wiring issue. A poorly manufactured USB cable can certainly cause a short circuit.

Fire investigators are also well versed in the forensic investigation of these scenes. If his expert analysis determined the cord was at fault, there's a good chance it was the cable and not the device.


> A short circuit is usually a wiring issue. A poorly manufactured USB cable can certainly cause a short circuit.

yes. also extremely high quality cables which have been damaged in the field can develop a short.

power supplies intended for use by users who cannot/will not inspect their cabling for damage (morally) should detect that the output is shorted and stop applying a voltage.

they're the only device in the "charger, cable, chargee" chain that can do anything about the situation.


It is a 5000 word article that talks about multiple classes of products, contextualizes the problem (regulations, competition, history, etc), and discusses possible mitigations.

It is a huge mischaracterization to suggest that this is a hit piece about "USB cords." That was one of half a dozen different electrical products they discussed, and only the jumping off point into a much larger discussion.


But the USB charger manufacturers could do better by shutting down the output should arcing/sparking be detected. We already have this for 120V AC supplies, and it should be doable for 5V DC.

https://en.wikipedia.org/wiki/Arc-fault_circuit_interrupter


>But the USB charger manufacturers could do better by shutting down the output should arcing/sparking be detected.

A poor connection at a plug can (or will?) cause a hot spot without arcing, and for a long thin cable a short may not even trigger the overcurrent protection of a high power charger.

Best to buy a cable with a decent 'AWG rating' eg 24AWG, either from a reputable brand like Tronsmart, or a specialist supplier, in the UK there's Kenable:

https://www.kenable.co.uk/en/




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: